We extracted clinical encounters for cases (including multiple episodes/patient) of acute gastroenteritis, nontyphoidal Salmonella infection, and Campylobacter infection recorded by MedicineInsight, a national primary healthcare database in Australia (https://www.nps.org.au/medicine-insight) during 2013–2018 and examined whether an antimicrobial drug was prescribed on the day of diagnosis (Appendix). Antimicrobial drugs were prescribed for 6.8% (6,652/98,496) of cases of acute gastroenteritis, including 35.7% (391/1,096) cases of nontyphoidal Salmonella infection and 54.1% (1,066/1,969) cases of Campylobacter infection.

Antimicrobial drug prescriptions for acute gastroenteritis increased with patient age (<10 years, 3.8%; >65 years, 13.7%) (Table 1). Antimicrobial drugs were more likely to be prescribed for those with than without the following: fever or no temperature measurement, a requested fecal sample test, underlying conditions, or a record of bacterial or parasitic infection. Antimicrobial drugs were less likely to be prescribed for those with a record of viral infection. Prescribing also differed by practice remoteness; prescribing was higher in practices in more remote areas than in cities. During the study period, the trend toward antimicrobial drug prescribing decreased from 7.8% to 5.8% (p<0.001). Similar findings were observed for children <10 years of age (Appendix Table 1).

The greatest reductions in antimicrobial drug prescriptions were found for those >65 years of age (2.8% absolute reduction from 13.4% to 10.6% (p = 0.049). The next greatest reductions were for those 30–49 years of age (2.4% absolute reduction from 8.3% to 5.9%; p = 0.006), 10–29 years (from 6.7% to 4.8%; p<0.001), and <10 years (from 4.8% to 3.0%; p = 0.03) (Figure 1).

For patients with nontyphoidal Salmonella infection (Appendix Table 2), prescriptions for antimicrobial drugs were more likely for those 30–49 than those <10 years of age (41.7% vs. 34.1%; p = 0.02) and in practices in outer regional or remote areas than in cities. Trend analysis of antimicrobial drug prescriptions for patients with nontyphoidal Salmonella infection suggested a significant reduction; absolute reduction was 11.4% (from 42.1% in 2013 to 30.7% in 2018; p = 0.01). For patients with Campylobacter infection (Appendix Table 3), antimicrobial drugs were more likely to be prescribed for female than male patients (56.8% vs. 51.7%; p = 0.02). We observed no significant reduction in antimicrobial drug prescriptions for patients with Campylobacter infection (55.8% to 57.1%; p = 0.81).

Of the 6,652 acute gastroenteritis cases for which antimicrobial drugs were prescribed, a reason was recorded for 42.9% (2,854/6,652), including 80.4% (2,295/2,854) for acute gastroenteritis, 1.1% (30/2,854) for other gastrointestinal illnesses, 5.7% (162/2,854) for respiratory tract infections, 1.8% (50/2,854) for urinary tract infections, and 11.1% (317/2,854) for other reasons. Of the 6,652 acute gastroenteritis cases for which antimicrobial drugs were prescribed, 7,159 prescriptions were written: 1 for 92.9% (6,179/6,652) of cases and >2 (range 2–5) for 7.1% (473/6,652). The predominant class of drug prescribed for acute gastroenteritis was nitroimidazoles (41.6% of total; Table 2), of which metronidazole accounted for the most prescriptions (24.7% of total; Appendix Table 4).

Prescriptions of cephalosporins, quinolones, and nitroimidazoles decreased significantly over the study period (Figure 2). The greatest reduction was for nitroimidazoles (absolute reduction from 3.9% to 2.3%; p = 0.001), followed by quinolones (1.3% to 0.8%; p = 0.02) and cephalosporins (0.7% to 0.5%; p = 0.049). However, prescriptions of macrolides increased significantly (0.6% to 1.0%; p = 0.01).

For the 391 cases of nontyphoidal Salmonella infection, a total of 418 prescriptions were written: 1 for 93.1% (364/391) and 2 for 6.9% (27/391). No dominant antimicrobial drugs were prescribed for patients with nontyphoidal Salmonella; most commonly prescribed were quinolones (30.4% of total; Table 2). For 1,066 cases of Campylobacter infection, 1,165 prescriptions were written: 1 for 91.0% (970/1,066) and >2 (range 2–4) for 9.0% (96/1,066). The predominant antimicrobial drugs prescribed for Campylobacter infections were macrolides (70.9% of total; Table 2), of which most were azithromycin (44.4% of total; Appendix Table 4).

In this large study of patient clinical encounters in general practices in Australia, we found that antimicrobial drugs were prescribed for 6.8% of all cases of acute gastroenteritis but for 35.7% of nontyphoidal Salmonella infections and 54.1% of Campylobacter infections. Over the 6-year study period, the absolute proportion of cases for which antimicrobial drugs were prescribed for acute gastroenteritis decreased by 2%.

Of the few studies reporting on how often antimicrobial drugs are prescribed for acute gastroenteritis, estimates range from 8.5% of 2,089 cases in a sentinel surveillance sample from primary care in Switzerland in 2014 (12) to 65% in a survey of 237 physicians in China in 2012 (13). Our results were most similar to the estimates reported from the Switzerland study, which also found that antimicrobial drugs were more likely to be prescribed for older patients and those with fever (12).

In Australia, treatment guidelines recommend that empirical prescription of antimicrobial drugs is of no benefit for acute gastroenteritis and is indicated only for patients with manifestations of severe disease, those who are immunocompromised, returned travelers of all ages, or children in whom systemic bacterial infection is suspected (7). Our results suggest that general practitioners are more likely to adhere to guidelines and that antimicrobial drugs are more likely to be prescribed for patients who are older, those with underlying conditions, and those with systemic symptoms (e.g., fever). However, the substantial numbers of patients without these indications for whom antimicrobial drugs were still prescribed suggests overuse of antimicrobial drugs for acute gastroenteritis.

Reassuringly, we did find reduced antimicrobial drug prescriptions for acute gastroenteritis during the 6-year study period. This finding is consistent with that of an earlier study that used the same dataset and found an overall reduction in the proportion of patients for whom systemic antimicrobial drugs were prescribed: from 31.7% in 2015 to 26% in 2017 (14). This reduction has been attributed to a series of antimicrobial stewardship programs implemented during 2009–2014, which included educational and advertising campaigns aimed at general practitioners and consumers (15). Our results suggest that these antimicrobial stewardship programs may have reduced antimicrobial drug prescriptions for acute gastroenteritis.

Given the estimated 1.1 million cases of acute gastroenteritis seen in general practices in Australia annually (3), we estimate that nationwide ≈74,000 antimicrobial drugs are prescribed for acute gastroenteritis every year. Because most of these drugs are probably unnecessary, our findings highlight the need for greater antimicrobial stewardship to support management of infectious gastroenteritis in primary care.